Senator Warren’s proposals seek to protect older technologies at the expense of consumers and workers.
Antitrust policy promises to be an important issue in the 2020 presidential election, and for good reason. Market power as measured by price-cost margins has been on the rise since the 1980s. Unreasonably high margins reduce output, causing higher prices for consumers and fewer jobs for labor.
One proposal would designate large platform-based companies such as Amazon as “platform utilities” and prohibit them from selling their own merchandise on the platform in competition with other sellers. This “structural separation” rule would apply to platforms that exceed $25 billion in annual revenue, including Amazon, Google, and Facebook. Smaller platforms would not be required to separate structurally but would be placed under a standard of “fair, reasonable, and nondiscriminatory dealing” with third-party merchants. These rules could be enforced by both government agencies and private parties.
Senator Warren’s second proposal would reverse “illegal and anti-competitive tech mergers,” focusing on large platform acquisitions of smaller firms. Her statement identifies Amazon and Whole Foods as such a merger, as well as Facebook’s acquisition of messaging company WhatsApp.
Both Warren proposals express concerns about privacy as well as competition. The proposals are populist and resemble, in two respects, the approach taken by the Trump Administration on climate change. First, the Warren proposals largely ignore the mainstream understanding of the problem. Second, although the proposals sound simple, they mask complex issues and are likely to produce unintended results. Structural antitrust remedies are easy to articulate and they have been a feature of heavy-handed antitrust policy since the early twentieth century. For example, a 1912 antitrust decree broke the Standard Oil monopoly into 34 smaller firms. But a few years later, the Federal Trade Commission was obliged to issue a report explaining why gasoline prices rose so dramatically after the breakup.
Why would a Warren Administration focus on large tech platforms? Although monopoly is a serious problem and margins have grown rapidly since the 1980s, the increases have mainly occurred in markets for manufactured products (such as cars), breakfast cereal, steel, airlines, household goods (such as batteries), and beer. By contrast, most of Google’s products are sold at a price of zero to consumers, although businesses pay for advertising. Amazon has very low margins in comparison to other retailers. Furthermore, the firms that the Warren proposal targets have high consumer satisfaction ratings and some of them, such as Amazon, have exhibited significant growth at the expense of traditional brick-and-mortar retail.
Senator Warren’s proposal is thus similar to the Trump Administration’s strategy of protecting coal at the expense of sustainable energy sources. Both strategies favor older technologies in danger of being displaced—fossil fuel companies in the case of the Trump Administration and traditional retail firms in the Warren proposal.
Who gains from Senator Warren’s first proposal to keep large platform companies from selling their own merchandise? Not consumers or labor, both of whom benefit from high output and low prices. Indeed, the text of the Warren proposal is largely indifferent to output or pricing—and may even lead to lower output and higher prices. The statement outlining the proposal justifies its prohibition by citing Progressive Era rules that required structural separation of public utilities from ordinary businesses. But utilities at that time, including railroads, were subject to price regulation. A price-regulated utility’s expansion into unregulated markets—for instance, when railroads acquired coal mines or hotels—created opportunities for manipulating the costs of its regulated business. By contrast, if it is not subject to price regulation, a company’s “vertical” integration of inputs and outputs can be good for competition. For example, oil refiners also sell gasoline in order to prevent price fixing by independent gasoline stations.
Firms also vertically integrate their distribution in order to eliminate price premiums that result from trademarks or entrenched brands. One very popular example is “AmazonBasics” house brand of consumer batteries, which are just as good as but much cheaper than branded batteries such as Energizer or Duracell. Energizer is owned by a holding company which also makes Rayovac and Eveready. Duracell is owned by Berkshire-Hathaway. Both are large publicly-traded firms with high price-cost margins.
House brands such as AmazonBasics allow customers to avoid paying high prices for the trademarks of other large companies. And when house brands are sold in competition with branded goods, as they are on Amazon, they also force name brands to cut their own prices.
But under Senator Warren’s proposal, Amazon could no longer sell AmazonBasics batteries in competition with these brands. The proposal faults house brands for copying the goods of others. To be sure, intellectual property laws are not always effective at preventing copying. Nevertheless, making a cheaper generic copy of another firm’s trademarked brand is not any kind of theft at all. Rather, it serves consumers by giving them the opportunity to avoid paying for a trademark or name that they do not want. The result is less monopoly, not more.
Senator Warren’s proposed exception for smaller platforms—allowing them to market both their goods and those of competitors—misunderstands the relationship between company size and market power. Amazon is a very large platform, but that is because it reportedly sells more than three billion products worldwide. For most individual products, Amazon’s market share is small. A more specialized vendor that sells only, say, computer printers might have much lower overall sales volume but a much larger footprint in that product—and consequently, greater market power. The distinction Senator Warren’s proposal makes based merely on company size makes no sense unless it is tied to a particular share of a product.
Senator Warren’s second proposal, which would keep large platforms from acquiring smaller firms, identifies a serious problem: large platforms are undertaking numerous acquisitions of smaller companies, some of which are or could become competitors. The problem is severe because predicting anticompetitive effects in individual cases is so difficult. On the one hand, such mergers can be procompetitive if they enable a large firm to improve its product offerings. On the other hand, small acquired firms might develop in the future into important competitors, but their acquisition prevents this from happening. One promising remedy to this problem is to limit larger companies’ technology acquisitions to nonexclusive licenses. That way, the firm can obtain any technology that the smaller firm has to offer, but it cannot acquire the right to exclude rivals from using the technology.
The Warren proposal, however, does not do this. Instead, it disapproves of mergers such as the one between Amazon and Whole Foods, which almost certainly benefit both consumers and labor. The Amazon-Whole Foods merger places rival grocery chains under pressure to innovate. For example, Walmart is now expanding its offerings to include delivery and other services.
Any time a merger or other practice reduces a firm’s costs or improves its products or services, it boosts competition by putting pressure on obsolete or less efficient rivals. But protecting these rivals should not be the purpose of the antitrust laws. Rather, the focus of antitrust laws should be on maximizing output, which benefits both consumers and workers.
In that respect, the Warren proposals are also notable for what they do not say. One stark deficiency is the lack of concern about high consumer prices, even though Senator Warren made her career promoting the protection of consumers. Nevertheless, her proposal ignores them and is more concerned about protecting inefficient, higher cost competitors.
Another problem is the lack of any concern about labor, whose wages have not come close to keeping pace with corporate earnings. Rather than looking out for competitors, antitrust policy should encourage maximum output, which helps both consumers and labor by providing more opportunities for jobs. If policymakers are truly concerned about workers, then they should urge that anticompetitive practices in labor markets receive greater attention from antitrust enforcers.
Not only are consumers and labor two of the Democratic Party’s most important and vulnerable constituencies, their protection is what should be driving antitrust policy for both parties.